ByteBrief
We're a portrait publication through and through. Turn your phone back and your briefing picks up right where you left it.
(We tried widescreen once. It wasn't us.)

OpenAI's custom Jalapeño chip, built with Broadcom, beat Nvidia's Blackwell-based GB300 on performance per watt in SemiAnalysis' InferenceX benchmarks. The 700-watt chip, tested with DeepSeek and Moonshot AI models, targets inference workloads and could cut costs by about 50%. Deployment begins by end of 2026.
Tap to vote and see what everyone thinks.
Summary by ByteBrief